Weighting Factors Design of Model Predictive Control for Three-Level Inverter-Fed PMSM Drives Using Multi-Objective Particle Swarm Optimization and Artificial Neural Network

Abstract

This article presents a novel strategy for designing weighting factors in the model predictive control (MPC) of a three-level inverter-fed permanent magnet synchronous motor (PMSM) by combining multi-objective particle swarm optimization (MOPSO) and artificial neural network (ANN). A cost function containing three weighting factors is designed to formulate a multi-objective optimization scheme. The MOPSO algorithm is employed to establish the Pareto front and calculate the optimal weighting factors for each operating condition. Additionally, an ANN is introduced for online tuning of the optimal weighting factors. Simulation and experimental results demonstrate that the proposed control scheme can solve the design problem of weighting factors, meet the real-time requirements of the inverter system, and exhibit excellent dynamic and steady performance.

Keywords

• PMSM
• MOPSO
• MPC
• weighting factors design